论文部分内容阅读
给出了一种能够综合运用先验知识和启发性知识的基于定性定量模型的动态过程故障诊断新方法。首先描述了核心的模糊定性建模和仿真技术,接着设计一种新的知识观测器,其中正常和已知故障状态的知识用模糊定性模型表示,将观测的行为与模糊定性仿真预测行为比较,将匹配的结果诊断输出,若找不到匹配结果,进行故障假设和生成相应模型,并将其预测行为和观测行为比较直到正确诊断。
A new method of dynamic process fault diagnosis based on qualitative and quantitative models that can comprehensively use prior knowledge and heuristic knowledge is given. Firstly, the core of fuzzy qualitative modeling and simulation technology is described. Then a new knowledge observer is designed, in which the knowledge of normal and known fault states is represented by fuzzy qualitative model, the observed behavior is compared with that of fuzzy qualitative simulation, Will match the results of diagnostic output, if can not find the matching result, the fault hypothesis and generate the corresponding model, and its predicted behavior and observation behavior comparison until the correct diagnosis.